Background

I would really want, to promise my customers "An approved Standart Security"; a real "security" on which they could rely on; and focus on my commercial products.

What I am doing, is not something I desire.

The history of "approved standards" is a proof to not to trust to so called "Security Authorities", "Experts" or "Cryptography Gods".

Do you want a proof? https://en.wikipedia.org/wiki/Transport_Layer_Security

This is a "proof"! A real proof to not to trust to "approved standarts"!

Not enough? Try this one: https://vikingvpn.com/blogs/security/visualizing-weak-encryption-experiments-with-aes and see with your own eyes what "protected" our secrecy for years!

SSL 3.0 doesn't include "any" secure encryption implementation. AES 256 CBC is not included in TLS 1.3 draft. All others are slow and again, untrustable considering the history.

I'd rather prefer to try to develop my own and improve it by seeing my mistakes and my bottlenecks with the transparent help from the community.

I am not a "crypto expert", I haven't mathematical background to prove that's "secure", as the "authorities" do!
But I believe in common sense and collaboration. I'd rather prefer to try to improve transparently a collaborative work, than relying on suspicious "Cryptoanalysis Gods" and "Authorities"!

If you think it is breakable, you're welcome, this is why it's in public domain. Please tell us "how"! Let's think together and improve it. We will all use this for our specific needs. And if we think carefully, we can adapt it to different requirements. Nobody prevents us from creating derivatives. This is freedom and the spirit of "open source".